A Big Data smart library recommender system for an educational institution
| DOI | https://doi.org/10.1108/LHT-06-2017-0131 |
| Published date | 17 September 2018 |
| Pages | 498-523 |
| Date | 17 September 2018 |
| Author | Aleksandar Simović |
A Big Data smart library
recommender system for an
educational institution
Aleksandar Simović
Department of E-Business, Faculty of Organizational Sciences,
University of Belgrade, Belgrade, Serbia
Abstract
Purpose –With the exponential growth of the amount of data, the most sophisticated systems of traditional
libraries are not able to fulfill the demands of modern business and user needs. The purpose of this paper is to
present the possibility of creating a Big Data smart library as an integral and enhanced part of the
educational system that will improve user service and increase motivation in the continuous learning process
through content-aware recommendations.
Design/methodology/approach –This paper presents an approach to the design of a Big Data system for
collecting, analyzing, processing and visualizing data from different sources to a smart library specifically
suitable for application in educational institutions.
Findings –As an integrated recommender system of the educational institution, the practical application of
Big Data smart library meets the user needs and assists in finding personalized content from several sources,
resulting in economic benefits for the institution and user long-term satisfaction.
Social implications –The need for continuous education alters business processes in libraries with
requirements to adopt new technologies, business demands, and interactions with users. To be able to engage
in a new era of business in the Big Data environment, librarians need to modernize their infrastructure for
data collection, data analysis, and data visualization.
Originality/value –A unique value of this paper is its perspective of the implementation of a Big Data
solution for smart libraries as a part of a continuous learning process, with the aim to improve the results of
library operations by integrating traditional systems with Big Data technology. The paper presents a Big
Data smart library system that has the potential to create new values and data-driven decisions by
incorporating multiple sources of differential data.
Keywords Libraries, Data analysis, Big Data, Data storage, Education, Recommender system
Paper type Technical paper
Introduction
Traditionally configured systems for data storage and analysis prevent libraries from
achieving competitive advantages. Over the past decade, many academic libraries have
struggled to shift the value and utility of collected data (Buckland, 2017). Library users have
switched to online scholar sources of information, and academic libraries have lost their
monopoly over the provision of scientific information (Chambers, 2013). Library users
retrieve new data necessary for learning, evaluate new theories, or discover a new addition
to knowledge. Each of these functions involves determining the specific knowledge,
professional literature and other learning materials that may not be available in the library
(Feisel and Rosa, 2005). Identifying and analyzing data beyond the library, through
campuses and external aggregations, can develop effective services and systems bringing
value to the wider institution (Showers, 2015). Large amounts of available data and the
implications of differential resources increase the complexity of their collection even further
(Showers, 2014). Library catalogs need to carry enough information about items and users
preferences to have the capacity to determine a potentially ideal result and to respond
adequately to the given query (Horstmann and Brase, 2016).
Nowadays, the cumulative increase in the volume of data from various sources
relational and non-relational (stored in local and in the cloud environment) has given rise
to the problem of providing an efficient library service to the users. According to
Library Hi Tech
Vol. 36 No. 3, 2018
pp. 498-523
© Emerald PublishingLimited
0737-8831
DOI 10.1108/LHT-06-2017-0131
Received 30 June 2017
Revised 30 October 2017
17 January 2018
14 March 2018
Accepted 14 March 2018
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0737-8831.htm
498
LHT
36,3
Rubin (2017), educational institutions are in a period of transition in how they deliver their
library services. New methods of content delivery allow e-libraries to substitute traditional
face-to-face librarian recommendations. Recommendation methods need to adapt/respond
to continually improving new technologies offering the possibility of new forms of
delivery just like the nature and structure of educational library catalogs might well
change so as to offer content of greater interest in line with the emergence of new
economic models. These challenging conditions make a smart library an inescapable part
of a modern library business system, which itself is an integral part of learning processes
and the educational institution.
The approach used in this work is based on the idea that the integration of Big Data
technologies into smart library data management ecosystem solves issues such as: how
large amounts of data from different sources can be collected and connected, integrated and
stored, and analyzed and visualized; and how to display the content of more interest to users
through a recommender system?
In order to improvethe process of meeting user needs inthe continuous educational cycle,
the proposed Big Data recommender system enables data integration from various sources
(e.g. Learning Management Systems (LMS), University online bookstore, Internet of Things
(IoT), data fromsocial media networks, and traditional library)into the smart library, making
the approach particularly suitable for application in educational institutions.
The main contributions of this paper are:
(1) Recommendationsystems were made basedon practical requirementsas personalized
e-services that haveapplication in different domains.Existing recommender systems
mainly focus on well-known approaches and they are reviewed in the next section.
This paper proposes an integrated recommendation system that complements
existing systems and provides a useful guide for librarians, practitioners, and
researchers in developing a Big Data smart librarymodel, and creating a new service
that will improve the educational process.
(2) This paper provides a framework for an efficient application of four independent
data sources into the Big Data ecosystem making the smart library an integrated
part of the educational continuum.
(3) For each data set, it effectively identifies the specific details and requirements for an
integrated recommendation with the aim of improving the results of library
operations by merging the traditional library and information systems (ISs) of an
educational institution with a Big Data framework. This will motivate and support
researchers and practitioners to promote the popularization of this approach.
(4) It particularly suggests possible further research of integrated recommendation
systems in the Big Data era with the proposition of developing a smart library
suitable for an educational institution.
The motivation behind this study is the lack of a comprehensive survey in the field of Big
Data smart library recommenders that approaches the issue from the perspective of an
educational institution as a foundation of knowledge creation and dissemination in society.
The paper is organizedas follows. First, it provides a reviewof the literature related to Big
Data technologies and recommender systems with a comparison of traditional systems and
other research fields in diverse domains where have found application. The next section
presents the BigData smart library model with the ultimate aim being the development of an
integrated recommender system suitable for an educationalinstitution with a detailed outline
of its implementation. That isfollowed by an evaluation of thesystem and its results. The next
section is a discussion. Finally, in the conclusion there is discussion which includes the
limitations, theoretical and practical implications of this research and outlines future work.
499
Big Data
To continue reading
Request your trial